MAP2K5 (MEK5) curation notes

UniProt Q13163 (MP2K5_HUMAN), 448 aa, HGNC:6845, 15q23. STE7-group dual-specificity MAP2K with an
N-terminal PB1 domain (~res 18-109) and C-terminal kinase domain.

Identity and family

• STE family MAP kinase kinase subfamily [file:human/MAP2K5/MAP2K5-uniprot.txt "Belongs to the protein kinase superfamily. STE Ser/Thr"].
• Cloned in human by Zhou et al. 1995 together with ERK5, which was found as a specific two-hybrid partner of MEK5 [PMID:7759517 "ERK5 was identified by a specific interaction with the MEK5 mutants S311A/T315A and K195M in the yeast two-hybrid system"; "ERK5 did not interact with MEK1 or MEK2"].
• Rat MEK5 cloned in parallel; does not phosphorylate ERK1/2/3, JNK or p38 PMID:7499418. Rat MEK5beta is "primarily cytosolic", MEK5alpha "particulate" PMID:7499418.

Molecular function: MAP2K for ERK5 only

• MEK5 is the dedicated and only known MAP2K for ERK5/MAPK7 [PMID:23382384 "MKK5 is the specific activator of extracellular regulated kinase 5 (ERK5)"; PMID:16260599 "In contrast to the other MAPK cascades, only one activator of ERK5, MEK5, has been cloned"].
• Phosphorylates Thr and Tyr of the ERK5 TEY motif PMID:14583600.
• In vitro reconstitution of MEKK3-MKK5-ERK5 module PMID:23382384.
• Constitutively active MEK5(D) selectively activates ERK5 in cells PMID:9384584.
• Deep research: ERK5 only established physiological substrate [file:human/MAP2K5/MAP2K5-deep-research-falcon.md "ERK5 remains the only firmly established direct physiological substrate of MEK5"].
• UniProt lists EC 2.7.12.2 and Ser, Thr and Tyr Rhea reactions; the Tyr and Ser reactions are generic parts of the dual-specificity MAP2K reaction, not separate functions. Better captured by GO:0004708.

Activation and scaffolding (PB1)

• Activated by MEKK2/MEKK3 through PB1-PB1 heterodimerisation [PMID:12912994 "The PB1 domains of MEKK2 and MEKK3 bind the PB1 domain of MEK5"; "the PB1 domain mediates the association of MEKK2 and MEKK3 with MEK5"].
• Activation-loop sites Ser311/Thr315 (S311D/T315D phosphomimetic activates ERK5) PMID:14583600.
• MEK5alpha N-terminus contains a unique ERK5 docking site (acidic cluster 61, 63-66) PMID:16260599.
• PB1 domain co-recruits activator and substrate: bivalent adaptor PMID:23382384.
• UniProt: "Acts as a scaffold for the formation of a ternary" MAP3K2/MAP3K3-MAP2K5-MAPK7 complex.
• Isoform note: MEK5beta (lacks 89 N-terminal residues incl. PB1) behaves as dominant negative in some studies PMID:14583600, though Seyfried et al. find both catalytically active PMID:16260599.

Biological roles

• Mouse Mek5 knockout: embryonic lethal ~E10.5 with cardiac defects, reduced proliferation/increased apoptosis; MEK5 required for ERK5 activation and MEF2 activity [PMID:15601854 "mek5(-/-) embryos die at approximately embryonic day 10.5 (E10.5)"; "MEK5 is required for mediating extracellular signal-regulated kinase 5 (ERK5) activation"].
• Endothelial shear stress: laminar flow induces KLF2 via MEK5/ERK5/MEF2; MEK5 activation "both required and sufficient" PMID:16341264.
• Constitutively active MEK5 in human endothelial cells: apoptosis resistance, decreased angiogenic/migratory/inflammatory potential via KLF4 PMID:20551324. This is the source (via rat RGD IMP, then ISS/IEA to human) of many BP annotations (IL-8, CXCL2, NF-kB, cytokine response, heterotypic adhesion, sprouting angiogenesis, apoptosis). These are downstream outputs of ERK5/KLF4 transcription, so non-core for MEK5.
• Cardiomyocyte hypertrophy (CT-1) needs MEK5-ERK5 PMID:15623435 -> source of rat positive regulation of cell growth.
• Cardiac anti-apoptosis via ERK5-CHIP-ICER PMID:20724525 -> source of rat "negative regulation of smooth muscle cell apoptotic process"; the abstract is about cardiomyocytes, so the cardiac muscle term (GO:0010667) fits better.
• Biliary epithelial proliferation in PCK rats, MEK5 siRNA inhibits PMID:15631999.
• IGF-2 pro-myogenic signalling in mouse C2 myoblasts PMID:19654213 -> source of IGF receptor signaling pathway.
• Mouse spindle IDA (PMID:15509711) is a broad phospho-antibody immunostaining survey of many kinases in M phase; weak for a specific MEK5 spindle location.
• TGF-beta increases MEK5 association with phospho-ERK5 and MEF2C in renal PTEC PMID:18588859; MEF2C is an ERK5 substrate, so MEF2C co-IP is likely indirect.

Interactions (GO:0005515 IPI rows)

• Informative: MAPK7 (substrate; NbExp=8), MAP3K2 (NbExp=7), MAP3K3 (NbExp=5) -> GO:0051019 / GO:0031435.
• HSP90AB1 (NbExp=4): kinase client of HSP90 (Taipale 2012) -> GO:0051879, consistent with BRAF review.
• SQSTM1 (PB1-PB1, by similarity in UniProt) - function unclear; LNX2, GRB2, ZNF620, ZNF426, GUCD1, PFDN5, DUSP21 - binary screens only; REMOVE generic binding.

Localization

• No human CC annotations other than an IEA spindle. Deep research: predominantly cytosolic; ERK5 (not MEK5) translocates to the nucleus. Rat MEK5beta cytosolic (PMID:7499418). Mouse has IEA cytoplasm (UniProt SL). Not added as NEW; cytoplasm used as core-function location context.

Decisions summary

• Core: GO:0004708 MAP kinase kinase activity; GO:0070375 ERK5 cascade; GO:0071499 cellular response to laminar fluid shear stress.
• Generic kinase MF terms (protein kinase / tyrosine kinase / serine kinase activity) -> MODIFY to GO:0004708 (consistent with MAP2K3).
• Signal transduction -> MODIFY to GO:0070375 ERK5 cascade.
• Did not propose NEW heart development / angiogenesis (necessity evidence from knockouts, not participation) nor MAP kinase scaffold activity (raised as question).

Audit of inherited review (2026-09-25)

• Re-checked every existing_annotation action and all supported_by quotes against cached files (all verbatim; validator clean).
• IPI partner mapping verified against GOA WITH/FROM (Q99759 MAP3K3, Q9Y2U5 MAP3K2, Q13164 MAPK7, P08238 HSP90AB1, Q13501 SQSTM1, Q06413 MEF2C, others from binary screens).
• Rat source of GO:0034392 confirmed via QuickGO as RGD IMP PMID:20724525. Open full text (PMC2992371) read: MEK5 experiments are all in neonatal rat cardiomyocytes / cardiac CA-MEK5alpha transgenic mice; the word "smooth" does not occur. MODIFY to GO:0010667 retained.
• Spindle (GO:0005819, IEA from mouse IDA PMID:15509711): abstract-only, does not name MEK5, full text not open -> changed MARK_AS_OVER_ANNOTATED to UNDECIDED per "do not overrule curators".
• Removed cytoplasm (GO:0005737) from core_functions locations: no human CC evidence, rat MEK5alpha is particulate and only MEK5beta is "primarily cytosolic" (PMID:7499418; PMID:16260599), so a location assertion is not well supported. Not proposed as NEW.